Nosocomial infections result in substantive increases in health care costs, length of hospitalization, morbidity and mortality. Most recently, attention has focused on the emergence of multi-antibiotic resistant pathogens as a cause of such infections; such microorganisms are becoming endemic in U.S. hospitals, further exacerbating already severe problems with nosocomial pathogens. Control of these pathogens will require a very sophisticated understanding of the underlying epidemiology of nosocomial infections. This will require a generation of clinical investigators with strong training in infectious diseases and epidemiology and biostatistics. The applicant, Anthony D. Harris, MD, MPH, board-certified infectious diseases, is currently an Assistant Professor in the Department of Epidemiology, University of Maryland School of Medicine. The proposed training is intended to provide him with a Ph.D. degree in Epidemiology, in the setting of a strong mentored research program. The proposed research focuses on methodologic issues critical to clinical investigations in infectious diseases and hospital epidemiology particularly relative to risk factor analyses of antibiotic-resistant organisms. The correct identification of potentially modifiable risk factors for antibiotic resistant pathogens is important. The study design used most often to identify risk factors is the case-control study. In the epidemiology literature, much has been written on the importance of the selection of controls in case-control studies. To date, many case-control studies performed to analyze risk factors for antibiotic resistant pathogens have used different control groups and thus yielded different results. The aims of the research are: l) To demonstrate that in case-control studies analyzing risk factors for antibiotic-resistant bacteria such as extended-spectrum lactamase (ESBL)-producing organisms, antibiotics identified as risk factors will differ depending upon the control group selected. 2) To identify modifiable risk factors for ESBL-producing organisms. 3) To quantify the importance of patient to patient transmission for ESBL-producing organisms. The first two aims will be studied by constructing a prospective, surveillance culture based prospective cohort of patients in intensive care units. These patients will be followed until they are either discharged from the ICU or develop an ESBL-producing organism. Nested case-control studies, with two different control groups, will then have multivariate risk factor analyses done. Variables that will be explored as possible risk factors include age, sex, antibiotics, and comorbid conditions. The results of the case-control studies will then be compared to a similar analysis done of the cohort. The comparison of the results of the case-control studies to the cohort will demonstrate the importance of control group selection. The second aim will be done by combining epidemiological relatedness and molecular relatedness through pulse-field gel electrophoresis of ESBL cultures to quantify the importance of patient to patient transmission.